DEVELOPMENT OF HIGHLY ACTIVE AND STABLE HYBRID CATHODE CATALYST

DEVELOPMENT OF HIGHLY ACTIVE AND STABLE HYBRID CATHODE CATALYST

3.1 INTRODUCTION

PEMFCs are attractive power sources of the future for variety of applications including portable electronics, stationary power, and electric vehicles. However, sluggish cathode kinetics, high Pt cost, and durability issues inhibit the use of PEMFCs for

automobile applications [83, 84]. A number of factors contribute to the performance degradation of PEMFCs including catalyst dissolution [12, 85-89], catalyst sintering [90, 91], membrane degradation [92-94], and carbon support corrosion [95-99].

One of the efforts in achieving increased catalytic activity is by alloying Pt with 3d transition metals to obtain high kinetic activity at 0.9 ViR-free for ORR [100-104]. In PEMFCs, Pt-alloys with various transition metals such as Cr, Co, Ni, etc. have been extensively studied and shown superior electrocatalytic activity for the ORR when compared to pure Pt [102-106]. The enhancement in kinetic activity over Pt by alloying Pt with transition metals is due to various factors including lowering of the Pt oxidation state [107], suppression of Pt oxide formation [107, 108], formation of a new electronic structure with higher Pt 5d orbital vacancies [100], decrease in the Pt-Pt interatomic distance and therefore a more favorable O2 adsorption [100], formation of a thin Pt skin

on the surface of the alloy core [109-111], and the altered electronic structures of the topmost Pt atoms [102, 103, 112].

Amongst the Pt-alloy catalysts, the PtCo catalyst has attracted much attention due to its high kinetic activity and stability in acidic environment [16, 105]. Paulus et al. studied the bulk compositions of 50 and 75 at.% Pt with Ni and Co as alloying elements [101]. In comparison to pure Pt, the results revealed a small activity enhancement of ca. 1.5 times for the 25 at. % Ni and Co catalysts, and a more significant enhancement by a factor of 2-3 for the 50 at.% Co. Huang et al. showed that PtCo alloy nanoparticles exhibit kinetic activity and specific activity enhancements by a factor of ~3.2 and ~2.2, respectively for the ORR when compared to pure Pt [113]. Antolini et al. reviewed the activity and stability of various Pt-alloy catalysts, and concluded that PtCr and PtCo are more stable than PtV, PtNi, and PtFe due to their high degree of alloying and particle size [105]. Jayasayee et al. studied the activity and durability of PtCo, PtNi, and PtCu in PEMFC cathodes as a function of alloying elements in a systematic manner [104]. They showed that the performance of PtCo and PtCu catalysts was found to be most attractive when compared to PtNi and Pt catalysts. Mani et al. investigated the kinetic activity of dealloyed PtCu, PtCo, and PtNi in PEMFCs [106]. They found that Pt-alloy with Co and Cu are more active than PtNi. Mass and specific activities of PtCo and PtCu were

enhanced by a factor of 3~4 times, compared to the commercial Pt/C catalyst. The durability of carbon-supported PtCo catalysts is the core advantage as

cathode catalysts in PEMFCs. Yu et al. studied the durability of Pt/C and PtCo/C cathode catalysts with continuous water fluxing on the cathode under a potential cycling test between 0.87 and 1.2 V vs. RHE [18]. The authors found that cobalt dissolution neither

detrimentally reduced the cell voltage nor dramatically affected the membrane

conductance. Cell performance enhancement by PtCo/C over Pt/C catalyst was sustained over 2400 cycles and the overall performance loss of the PtCo/C membrane electrode assemblies (MEAs) was less than that of the Pt/C MEA. Arico et al. reported the performance and durability of carbon-supported PtCo under high temperature (110- 130 °C) operation in PEMFCs [114]. A potential cycling test at 130 °C in a pressurized PEMFC showed a better stability for the PtCo alloy than pure Pt/C. Furthermore, better performance was obtained at high temperatures for the pre-leached PtCo/C than the Pt/C catalyst. They observed that the amount of Pt oxides on the outermost atomic layers was much smaller in PtCo than in Pt catalyst. These characteristics appeared to influence catalysts’ performance and durability. Stassi et al. investigated the effect of thermal treatment on the structure and surface composition of PtCo catalysts during accelerated stress test (AST) [115]. They reported that different thermal treatments caused significant structural and morphological modifications in the PtCo catalysts. Yu et al. studied the cycling stability of dealloyed PtCo3 and PtCu3 catalysts between 0.6 and 1.0 V (vs. RHE) for up to 30,000 cycles [116]. In situ X-ray absorption spectroscopy (XAS) analysis showed stronger bulk Pt-Pt compressive strains and higher bulk d-band vacancies for the dealloyed PtCu3 than the dealloyed PtCo3 which was correlated to the higher initial activity of dealloyed PtCu3. MEA tests showed poor durability towards voltage cycling for the dealloyed PtCu3 catalyst when compared to dealloyed PtCo3 catalyst due to Cu plating on the anode.

under a reducing atmosphere and acid leaching procedures [106, 116, 117]. Since excess transition metal salts are used for the catalyst synthesis, the leaching is carried out in strong acids for prolonged time which may be detrimental to the support stability when the cathode experiences high potentials during startup/shutdown cycles. In our previous studies, we reported a novel method of preparing Co-doped Pt catalysts on CCC supports [85, 118]. In the present study, Co was initially doped into the CB at high temperature using metal-catalyzed pyrolysis which was used as a transition metal source for the formation of Co-doped Pt. The Co-doped carbon prepared in this manner was used as a support to deposit Pt nanoparticles (Pt/CCC catalyst synthesis). The Pt supported on Co- doped carbon was heat-treated under reducing atmosphere to obtain Co-doped Pt catalyst with controlled particles size. During heat-treatment, Co, which is doped within the carbon, diffuses to the surface and forms Co-doped Pt catalyst with a core-shell structure. The kinetic activity and durability of kinetic activity of Co-doped Pt prepared by the novel approach were examined and compared with those of commercial PtCo/C as well as state-of-the-art Pt/C catalyst [119, 120].

3.2 Experimental

3.2.1 PREPARATION OF SUPPORT AND CATALYST

The CCC support was prepared using the procedure developed at the University of South Carolina [51-53, 61, 65]. In brief, as-received CB (Ketjen Black EC-300J) was oxidized with 9.8 M HNO3 solution at 85 °C for 9 h under refluxing conditions. After filtering, the oxidized CB was washed with DI water several times and dried under vacuum at 80 °C for 12 h. A desired amount of Co(NO3)2 and ethylene diamine, used as

Co and N precursors, respectively, were mixed with the oxidized CB in 200 ml IPA. The mol ratio of Co and N precursors was maintained at 1:9. The mixture was reflexed for 3 h at 85 °C under vigorous stirring, followed by drying under vacuum at 80 °C. The

resultant powder was subjected to heat-treatment under inert atmosphere at 800 °C for 1 h followed by leaching in 0.5 M H2SO4 at 80 °C for 3 h to remove excess Co. The final product is denoted as CCC. The CCC was e non-covalently activated by the 1-

pyrenecarboxylic acid (PCA) before the Pt deposition [121, 122].

Pt deposition was accomplished by a polyol reduction method for the preparation of 30% Pt/CCC catalyst. First, the CCC support was dispersed in 25 ml of ethyleneglycol in a sonication bath (Branson ultrasonic cleaner). A desired amount of PtCl4 was added and the pH was adjusted to 11 by the addition of 0.1 M NaOH solution. The resulting solution was refluxed at 160 °C for 3 h and allowed to cool to room temperature. Then, the solution was filtered, washed with DI water, and dried at 160 °C for 1 h. Prior to heat- treatment, the Pt/CCC was subjected to a protective coating procedure using polyaniline. Oxidative polymerization of aniline sulfate was carried out at room temperature using ammonium peroxysulfate as the oxidizing agent [123, 124]. The polyaniline-coated Pt/CCC was placed in an alumina crucible and heat-treated at 700 -900 °C for 2 h in a tubular furnace under 5% H2 (balance N2) atmosphere. The catalyst thus prepared is denoted as Co-doped Pt/CCC.

For comparison, the PtCo catalyst was prepared by the conventional

impregnation method with the same ratio of Pt to Co as in the Co-doped Pt/CCC. Briefly, the 46 % Pt/C (TEC10E50E, Tanaka Kikinzoku Kogyo K.K, Japan) is mixed with

homogeneous slurry. The resultant slurry was dried in an oven for 12 h under vacuum followed by heat-treatment at 800 and 900 °C for 2 h under 5% H2 (balance N2) atmosphere. The catalysts thus obtained were denoted as PtCo/C-Imp-800 and PtCo/C- Imp-900, respectively.

3.2.2 PHYSICAL CHARACTERIZATION

The nitrogen adsorption/desorption isotherms were obtained at −196 °C using a Quantachrome NOVA 2000 BET analyzer. Specific surface area was determined by a multipoint BET analysis. PSD curves were calculated by the BJH method using the adsorption/desorption branch. XRD analysis was performed using a Rigaku D/Max 2500 V/ PC with a Cu Kα radiation. A tube voltage of 30 kV and a current of 15 mA were used during the scan. To estimate the particle size of samples, we employed the following Scherrer equation [66]: 10 cos k D B    [3.1]

where D is the crystallite size in nm, k is a coefficient (0.9), λ is the wavelength of X-ray (1.5404 Å), B is the line broadening at half the maximum intensity in radians, and θ is the angle at the position of the maximum peak known as Bragg angle. Raman spectroscopy was used to evaluate the degree of graphitization of the carbon supports using HORIBA "LABRAM 1B” (He-Ne 20mW laser, wave length 632.817 nm). Inductively coupled plasma atomic emission spectroscopy (ICP-AES, Perkin Elmer) analysis was used to determine the composition of the catalysts. HR-TEM was used to study the morphology and particles size distribution of the catalysts using Hitachi 9500 HR-TEM operated at

300 kV accelerating voltage. XRF (Fischer XDAL) was used to determine PtCo composition in the catalyst and Pt loading in the catalyst coated membrane.

3.2.3 RDE MEASUREMENTS

A glassy carbon disk electrode (0.247 cm2) was acted as a working electrode. The Ag/AgCl electrode and platinum mesh were used as a reference and counter electrodes, respectively. All electrode potentials reported here were converted into the RHE. In a typical RDE experiment, 8 mg of CCC and CB was ultrasonicated in 1 ml of IPA. 15 ul of the ink (0.12 mg cm-2) was deposited on the glassy carbon electrode. For the Pt/CCC and Pt/CB, the catalyst was mixed with absolute ethanol and DI water ultrasonically. The catalyst ink was deposited onto the glassy carbon electrode.

RDE tests were performed in 0.1 M HClO4 solution as an electrolyte at room

temperature using a Pine bi-potentiostat (Model AFCBP1). The CV was swept at a scan

rate of 50 mV s-1 _{from 0.005 to 1.0 V in deaerated electrolyte under N}

2 atmosphere. LSV

measurements were conducted at a scan rate of 5 mV s-1 _{in O}

2–saturated electrolyte by

sweeping potential between 0.2 and 1.1 V anodically. The LSV curves presented in this

work are properly corrected using the background capacitance current that is measured in

the N2 atmosphere at a scan rate of 5 mV s-1.

3.2.4 MEA FABRICATION AND ELECTROCHEMICAL MEASUREMENT

For the MEA fabrication, the in-house synthesized catalysts were employed as the cathode catalyst while commercial 46% Pt/C was used as a catalyst for the anode.

IPA (1.8 ml), Nafion® ionomer (5% solution, Alfa Aesar), and DI water (0.2 ml). The ionomer content was 30% and 20% in the anode and cathode inks, respectively. The catalyst inks were sprayed directly on the Nafion® 212 membrane covering an active area of 25 cm2. The Pt loading on the anode and cathode electrodes is kept at 0.1 and 0.15 mg cm−2, respectively. The catalyst coated membrane was then hot pressed at 140 °C using a pressure of 20 kg cm−2 for 6 min. in between the gas diffusion layers (Sigracet GDL 10BC, SGL) and Teflon gaskets to prepare the MEA for the performance evaluation studies in fuel cell.

Initially, the MEA was activated under a supply of H2 and O2 at 80 °C to the anode and cathode compartments, respectively with a flow rate of 750 sccm and 100% RH. After MEA activation, the initial kinetic activity at 0.9 ViR-free was evaluated under H2/O2 (2/9.5 stoic.) at 80 °C, 100% RH, and 150 kPaabs. back pressure. The

electrochemical surface area (ECSA) was estimated using CVexperiments carried out between 0.05 and 0.6 V (vs. RHE) at 80 °C under fully humidified H2 and N2 supply to

the anode and the cathode, respectively. The mass activity measurements were performed

using the AST protocol suggested by U.S DRIVE Fuel Cell Tech Team [125]. During

AST, 200 sccm H2 and 75 sccm N2 were supplied to the anode and cathode, respectively and the potential was swept between 0.6 and 1.0 V (vs. RHE) at 50 mV s−1 in a triangle profile for up to 30,000 cycles. The fuel cell polarization was conducted using a fully automated fuel cell test station (Scribner Associates Inc., model 850e) at 80 °C. The mass

activity and ECSA measurements were performed after 0, 1000, 5000, 10,000, 20,000, and 30,000 cycles. The cell potential loss at 800 mA cm−2 _{was used as one of the criteria} to evaluate the catalyst performance. For comparison purposes, MEAs with commercial

PtCo/C (TEC36EA52, 46.8% Pt and 6.7% Co, Tanaka Kikinzoku Kogyo K.K, Japan) and 46% Pt/C (TEC10E50E, Tanaka Kikinzoku Kogyo K.K, Japan) as cathode catalysts were also prepared and evaluated under the same experimental conditions.

3.3 RESULTS AND DISCUSSION 3.3.1 CCC SUPPORT SYNTHESIS

Figure 3.1 illustrates the schematic diagram of the approach used to synthesize CCC and Co-doped Pt/CCC. Surface modification on the carbon support introduces oxygen and nitrogen groups on the surface [51, 65]. The metal-catalyzed pyrolysis increases the carbon graphitization degree at high temperatures (800~900 °C), introduces 7-15% Co in the carbon matrix, and incorporates nonmetallic (nitrogen-containing) active sites on the carbon surface. Next, the chemical leaching removes excess Co, and Co particles encapsulated with thin carbon layer is remained in the support which is used to dope Pt for preparing Co-doped Pt/CCC catalyst. Modified polyol process was used for uniform platinum deposition [118]. In the subsequent annealing step, the Co encapsulated within the CCC support diffuses to the surface to form Co-doped Pt catalyst in the

presence of polymer protective coating [118]. The heat-treatment process was optimized to control the particle size between 3-5 nm, resulting in compressive Pt-lattice catalyst having Pt-shell/doped metal core structure.

Figure 3.2 (a) and (b) show the nitrogen adsorption-desorption isotherms and BJH PSD curves of CCC and CB. The specific surface areas of CCC and CB are 398 and 826 m2 _g−1_{, respectively. The CCC exhibits characteristic Type IV adsorption/desorption}

Figure 3.1 Schematic diagram for CCC support and Co-doped Pt/CCC catalyst synthesis. Pt shell Co or Pt-Co core Carbon with active sites Graphitization CCC (Co-doped) Pt/CCC Co-doped Pt/CCC surface modification Cobalt- catalyzed

pyrolysis Chemical_{leaching deposition}Pt

Controlled pyrolysis

Figure 3.2 (a) N2 adsorption/desorption isotherms and (b) BJH pore-size distribution curves obtained from the adsorption branch of CCC and CB. The inset in (b) compares the PSD in the range 0-10 nm.

(a)

isotherm behavior according to the IUPAC classification indicating its mesoporous nature [67]. The isotherms show hysteresis loop with sharp adsorption and desorption branches over a relative pressure range of 0.4-0.8. The nitrogen uptake is observed when (P/P0) ratio is 0.94-1.0, which indicates the presence of mesopores [67]. The total pore volume was reduced from 0.846 to 0.688 ml g−1. As shown in Figure 3.2 (b) inset, after the metal- catalyzed pyrolysis the peak pore diameter is ca. 4 nm.

Figure 3.3 (a) presents XRD patterns of the CCC and CB. Generally the

characteristic diffraction peaks of (002) and (101) planes for carbon are found at ca. 26 and 43°. The diffraction peaks of CCC are sharper with increased intensity and shift to more positive angles. Consequently, the interlayer spacing of CCC based on (002) plane decreases to 0.3456 nm, while that of CB is 0.3615 nm. The results indicated that the carbon surface of CCC has been partially graphitized during metal-catalyzed pyrolysis. Furthermore, the CCC shows characteristic diffraction peaks at 44.2, 51.5, and 75.8° which correspond to the (111), (200), and (220) planes of FCC structure of Co metal particle (PDF#97-007-6632), respectively. The XRD results confirm the presence of Co metal after acid-leaching at 80 °C. Additionally, Figure 3.3 (b) reveals the Raman spectra for CCC and CB. Both CCC and CB show the D band and G band at approximately 1350 and 1580 cm−1, respectively. The D band originates from structural defects and disorder- induced features on carbon, while the G band corresponds to the stretching vibration mode of graphite crystals [70, 126]. Relative ratio of D band to the G band (ID/IG) for CCC and CB is estimated to be 2.42 and 2.60, respectively, indicating that CCC is more graphitized than CB.

Figure 3.3 Comparison of (a) XRD patterns and (b) Raman spectra of CCC and CB.

(a)

The HR-TEM images of CCC (after acid leaching) and CB are shown in Figure 3.4 (a) and (b), respectively. The apparent difference between them is the presence of Co particles encapsulated by carbon shells in the CCC support since the Co particles present on the surface are removed during acid leaching. Nanostructured fibers or tubes of graphitic carbon are also formed as a result of pyrolysis in the presence of Co metal [55, 56] while CB showed amorphous morphology as shown in Figure 3.4 (b). ICP-AES analysis of CCC indicated a cobalt content of ~13 wt% in the CCC synthesized at 800 °C.

The results of XPS analysis performed on CCC and CB supports are presented in Figure 3.5 (a) and (b), respectively. Figure 3.5 (a) shows the survey scans for CCC and CB. Only XPS spectrum of CCC, as shown in Figure 3.5 (b), exhibits a broad peak around 398.9 eV which corresponds to the nitrogen atom. The nitrogen peak shown in Figure 3.5 (b) for CCC can be deconvoluted into four major peaks corresponding to pyridinic, pyrrolic and/or pyridone, quaternary, and pyridinic-N+-O− (oxidized nitrogen). The peak at 398.4 eV accounts for the presence of pyridinic-N whereas the peak at 400.3 eV corresponds to the pyrrolic-N and/or pyridine-N. The peaks at 401.1 and 403.4 eV are ascribed to the presence of quaternary-N and pyridinic-N+-O−, respectively. Relative percentages of pyridinic-N and pyrrolic-N and/or pyridine-N are 41 and 38.5 % of total nitrogen, respectively. Quaternary-N accounts for 6.6 % while pyridinic-N+-O− occupies 13.9 %. It is well-known that pyridinic-N situated on the edge of the graphite planes promotes ORR by donating one p-electron to the aromatic π system [55, 68]. Moreover, previous studies report that the quaternary-N plays a role as stable ORR active sites [55, 56, 61, 70]. Results for the deconvolution of the N 1s spectra are summarized in Table 3.1.

Figure 3.4 HR-TEM images of (a) CCC and (b) CB.

(a)

(b)

Figure 3.5 (a) XPS survey scans of CCC and CB and (b) deconvoluted N1s XPS spectra of CCC.

(a)

Table 3.1 Characteristics of CCC obtained from XPS N1s peak BE [eV] FWHM Relative intensity [%] Pyridinic- N 398.4 1.736 41.0 Pyrrolic N and or pyridone-N 400.3 2.254 38.5 Quaternary-N 401.1 2.000 6.6 Pyridinic-N+-O- 403.4 4.331 13.9

3.3.2 ELECTROCHEMICAL CHARACTERIZATION OF CCC

Figure 3.6 exhibits the electrochemical properties of CCC in a half-cell employing

0.1 M HClO4 as the electrolyte. Figure 3.6 (a) represents the CV curves of CCC and CB

at room temperature. CCC shows the quinone/hydroquinone redox coupling peaks at ca.

0.6 V; as the potential increases, the oxidation current of CB also increases while that of

CCC remains almost constant. The low oxidation current may be attributed to the